Rural Hertfordshire in the Iron Age and Roman Period
by Ilid E. Anthony

The period from 500 BC to 500 AD is one of intensive development in the agricultural activity of south-eastern Britain. It becomes evident with the extension of archaeological excavation on both Iron Age and Roman sites that there was far greater continuity of methods than had formerly been suggested by the studies of archaeologists. Sites were studied in isolation; villas were simply buildings where mosaics might be found; Romano-British towns and forts attracted the attention of antiquarians and made them belittle the clearing of trees and the draining of swamps which made more cultivation possible.

Within the county of Hertfordshire, a purely arbitrary area for the purposes of this paper, two villa sites have been completely explored according to modern standards. It has been possible to study more of the type of agricultural activity which was used. Iron Age occupation was detected on both sites. Several of the Iron Age sites within the area have also been excavated and from all these studies a regional survey, though still incomplete, can be attempted.

The folded chalk hills form part of the range which fills the western and northern part of the county. In this area most of the prehistoric finds indicate much trafficking and settlement on the Chiltern range. The earliest Iron Age settlements are also here. In the southern half of the county the clay and the gravel similar to that of the Thames Basin is found. Bronze Age hoards have been recorded in the south-west. But the distribution map for pre-Iron Age sites is incomplete and apt at present to give a false picture. Considerable research in certain areas distorts the scene.

But it is a fact that the many river valleys of Hertfordshire were the main lines of penetration from early times; Iron Age finds as well as the Roman homesteads are plotted along their courses. The first group is that of the Colne valley and its tributaries the Chess, Gade, Misbourne and Ver. The second group is that of the Lea and its many side streams, the Mimram, Bean, Rib, Ash and Stort, which are in the centre of the county. The only rivers which flow to the north are the Hiz, Ivel and Rhee.

The Iron Age in Britain as in every other part of north-west Europe was a period of migrations; tribes driven out of their homes in eastern Europe by invading hordes from central Asia penetrated to the Atlantic sea-board from the north near the Rhine estuary in France and in Spain. From several points on this sea-board various invasions came across to Britain. Their greatest need was to find land on which to settle, but their entrance into hostile territory caused them to be warlike and to choose strategic sites for their settlements and to protect these with high earth ramparts and outer ditches. The ramparts were crowned with a wooden palisade and the entrances were flanked with guard-chambers; timber was also used to revet the rampart and prevent it from collapsing into the ditches. In areas where stone was more abundant, as in the Welsh Marches or Cornwall, it replaced the timber. Similarly with the huts which were normally built behind the ramparts where they could be protected from the winds as well as from attackers, in other areas they were of stone but in Hertfordshire they were of wattle and daub. Circular in plan, they would have had one stout post in the centre of the hut as the main support for the roof, which would undoubtedly have been of thatch. Circular hut plans were recovered at Arbury Banks near Ashwell during exploratory digging in the early part of this century; they were four to five feet in diameter.

The elongated horseshoe plan of Arbury Banks as seen today is not a true picture of the fort, since levelling of the rampart has taken place in modern times. Its defence system followed the contours of the hill, as do the ramparts at Ravensburgh Castle at Hexton and Ivinghoe Beacon near Aldbury. In the former the plan gives a rectangular area with the entrance on the west side. The entrance at Arbury has been destroyed, while there are clear traces of incurving ramparts to mark the entrance at Ivinghoe.

Incurving ramparts protected the guard towers on either side of an entrance while at Ravensburgh another form of protection may have been devised, namely the formation of a passage between two ramparts formed by one being extended for a short distance beyond the other. The cart track leading up to the fort, though much damaged, may have been on this pattern. These Iron Age A forts, as they are called by British archaeologists, were constructed in the fifth century BC when the earliest iron-using tribes crossed over to south-eastern Britain from northern France, bringing not only the knowledge of how to make iron weapons and tools but also the influences of the Hallstatt culture into the island. This is seen in the specially elaborate swords, brooches and horse-trappings that are associated with the culture. Examples of these have been found scattered in the valley of the southern half of the county as well as in the hill-forts on the chalk ridge from Ivinghoe, which can be mentioned with Crab Hill near Gerrard's Cross in Buckinghamshire, while Maiden Bower at Dunstable is in the vicinity of Ravensburgh Castle. Wilbury Hill has been extensively excavated and some of the pottery of the earliest levels appears to be of the Iron Age A type. High-shouldered situlate jars of gritty black brown were made from local clay.

Occupation of these early forts may have been spasmodic when the tribes had finally established themselves in the district. It is quite obvious that the herds of sheep, pigs and goats would have been allowed to graze outside the stronghold when no danger threatened. Since our knowledge of the extent of the occupation is so small, it can only be suggested that by the time that the Belgic invasions were being a threat, isolated farmhouses, on the pattern of Little Woodbury near Salisbury, had been established within the region. The exploration of the areas around Wheathampstead, a site from which much pottery has already been obtained, might reveal such a homestead. The fort which today stands south east of the modern village has been damaged by modern agricultural activity, but originally it was a hundred acres in area and its rampart was about thirty feet high. Westland Grove near Little Hadham and Wallbury near Broxbourne are others that could be examined. Ivinghoe Beacon appears to have an annex added to its original rampart which may indicate a renewal of occupation.

Forts in Sussex and in the western area of Britain were re-occupied and the defences recommissioned in times of danger -danger from the Belgae or Iron Age C or later against the Roman invasion. The Iron Age B were the tribes who settled in the south and west of Britain; they do not penetrate into this area. Therefore the remaining Iron Age sites within Hertfordshire are of the Belgic phase and were established during the first century BC. In the beginning there was the need to protect homes with a defence system, but by the mid first century there seem to have been considerable dwellings both on the gravel terraces of the rivers and the high chalk plateaus. There must have been trackways which connected the various settlement areas. There were dykes or long sinuous defence systems between the valleys, as between the Ver and Lea valleys, which were protected by the massive earthwork which today has various names - the Devil's Dyke at Gorhambury, Beech Bottom near the St Albans and Harpenden road. The road to Wheathampstead through Sandridge follows the continuation until it is obliterated before reaching the Wheathampstead site.

Since our knowledge of Ivinghoe is based mainly on field-work and the dating is only vaguely as yet noted in pottery picked up on the site, we cannot declare it to have been a Belgic stronghold. But the fort at Wilbury was reoccupied and elaborated. The Aubreys near Redbourn with their double ditch system may be a Belgic site, though the trench dug there in the thirties did not secure dateable occupation material. Its defences are more elaborate than those of Prae Wood, a site, though better known due to its exploration by the Wheelers in 1933-34, which has defences which seem insufficient when compared with the Aubreys.

Not only were the pottery forms used by the Prae Wood inhabitants more elaborate than those at Wheathampstead, which was the earlier settlement, but they showed foreign influences. Trading contacts with the continent were greater, since these people had gold, silver and bronze coins which they appear to have minted in the valley of the Ver, though their dwellings and herd enclosures were on the chalk plateau above the river and the site where the Roman city was later to stand. These rough imitations of the coins used by the Greeks and Romans have sometimes the names of the tribal leaders and sometimes the name of the tribal capital, as when 'Ver' or 'Verlucio' are found on these coins. With these or with their exports of corn or hunting dogs these people traded with the Gaulish merchants for amphorae filled with wine or oil, or with the exporters of Gallo-Roman pottery for the highly burnished platters or thin white or buff beakers, or from Italy itself the beautiful red Arretine dishes. All these may have been the possessions of the rich farmers but they influenced the local potters, who with the arrival of the Belgic people had for the first time in Britain the potter's wheel. The better technique and the more efficient firing of these potters on the eve of the Roman conquest is evident when the pottery of Prae Wood is compared with that of Wheathampstead.

In ploughing there was also an advance since the Belgae introduced the heavier plough into Britain. Therefore already before the landing of the Roman soldiers the trees of the river valleys were being gradually cleared and heavier clay was being turned for the first time. Park Street, two miles south of Verulamium, had a farmer who lived in a rectangular timber house and continued to live in that house until 75 A D, carrying on from generation to generation on the same site though the local tribes had submitted to the Roman emperor and the administration was henceforth to come from the tribal centre at Verulamium. It was the attraction of living in a Roman town where traffic passed through the valley that brought some of the inhabitants from Prae Wood and elsewhere to the new town. Even though some may have become shopkeepers and others been forced to become slaves through economic conditions, the link with the countryside was not broken. Herds would still be grazed outside the town and brought in at night, and people would have to maintain their own plots of ground outside the town to grow the corn which from then on had to be paid as part of their taxes to the Roman provincial government. As Strabo declared that corn was one of the exports of Britain before the conquest, it would have to be increased because of the demands of the government and the new rearrangement of the social order.

An examination of the records shows in the Colne river complex a series of villa sites along the valleys. At Moor Park between Rickmansworth and Watford a Roman building was excavated in recent years by the boys of the Merchant Taylors' School. This appeared to be of the general date from the first century to the end of the fourth. Similarly the site at Hamper Mill, a report on which is due to appear in the next volume of the St Albans Society Transactions, was a site in the same valley. They have a similar character. A building was recorded at Sarratt including one room which appears to have an apsidal end. Roman pottery has been found at Garston but insufficient to indicate a dwelling. A little distance from the river itself Roman occupation material has been found at Aldenham and Bushey, but the records are incomplete. At Munden, on the estate between Garston and Bricket Wood, a tumulus of a very elaborate character was recorded as containing an elaborate urn. Beneath the lawn of Munden house the hypocaust system of a villa showing four structural alterations has been noted - the occupation material was of the first and second century but the house was clearly an elaborate one.

At the junction of the Ver and the Colne another Roman site had been detected by Dr Norman Davy during the last war. In 1960 the Watford Archaeological Society have explored the site still further and have found an elaborately heated room and a deep plunge-bath lined with opus signinum. The main series of rooms, as in Munden, have yet to be detected and properly dated. Should all these sites be fully cleared and the date of their bath wing be ascertained it might be that the scheme which has been found both at Park Street and Lockleys, Welwyn, might have applied to them: namely that the villas begin with a series of rooms which appear to be plain living rooms. Only after a considerable period do they have their more sophisticated features such as the bath-wings added. The villa site at Park Street only two miles south of Verulamium was threatened and had to be completely excavated. Though only a simple dwelling until the late third century, when it received its hypocaust room, there was no break of continuity between the Iron Age and the Roman period - neolithic pottery even had come out of pits below the first timber structure. As in all farms storage space was needed and and a cellar was constructed.

Another example of a cellar was found at the villa site at Gorhambury. The Roman house stands within a mile north west of Verulamium and less than a mile from the Roman route now called Watling Street, which passed through the valley of the Ver. The cellar, a feature found at Park Street and possibly Sarratt, was the first part of the building to be uncovered. This had an elaborate apsidal end and an entrance ramp 5 feet 6 inches wide. Holes for shelving supports in the walls and marks in the plaster of the walls confirm its use as a store. The limestone block in the floor must have been the stantion for more shelving, perhaps for heavy amphorae. During the excavation of this cellar two levels of burning were found on the floor. In one of two pits which were in this floor was a coin of Hadrian (AD 117-132). In the apse two mid-second-century coarse pottery bowls were buried. This evidence confirms that the cellar was in existence in this century at least. The room above must have been lavishly decorated with painted plaster and with stucco figures since fragments were found in a tip line in the cellar. But the whole feature was drastically altered, if not completely closed, in the fourth century since another building had been erected over it. This was part of a fourth-century bath-wing consisting of two heated rooms and a plunge bath. The pottery from this wing and from the verandah which stood on the western side includes painted coarse ware and imitation Samian. This is the same development as in the other sites, though the Belgic occupation has not yet been discovered at the latter. First-century walls have survived in the form of chalk foundations, but the extent of the building of that period is not known. Air photographs have shown other buildings near the corridor house which could be farm buildings, as in the Ditchley villa near Oxford.

The nearest villa to Verulamium on the western side was a site which produced pottery ranging from the early second century to the fourth at Breakspear Farm. This site was discovered when excavating operations for M1 were in progress. The site was examined and the rectangular plan of the building observed, but since the bulldozer had leveled it nothing more could be done.

The north-west area of the county from Tring to Hemel Hempstead is occupied by the valleys of the Misbourne and the Gade. Roman sites have been detected from Moneybury hill, near Aldbury, to Nash Mills. Roman coins of the first century have been found at Northchurch and Berkhamsted, but a villa was found at Boxmoor with five rooms in one block as in the main wing at Park Street. This villa was found in the last century by Sir John Evans.

Further Romanisation north of Gorhambury has been noted at Harpenden, where a temple and cemetery suggest a considerable rural settlement. In the valley of the river Mimram two villas stand near to each other, the villa at the Manor House and the fully excavated villa at Lockleys. Roman material has been found at Welwyn Garden City at Bessemer Road. Still more Roman tiles and other occupation material has been found recently during road-widening operations on the A1. These may be of further buildings connected with the villa at Lockleys but they could prove to be an additional site.

In the Lea valley a villa has always been accredited to Amwell, where traces of a kiln are also recorded.

In the valley of the Hiz there is a villa at Great Wymondly and in the town of Hitchin itself Roman occupation material has been found. From Hitchin to Letchworth along the Roman road which ran parallel with the Icknield Way Roman occupation and burials have been recorded. An extensive cemetery at Baldock again suggests a considerable settlement but its actual site has not been recorded. Roman pottery and coins sometimes occur on Iron Age sites, e.g. Ravensburgh Castle and Arbury Banks, and also at isolated Iron Age sites such as Barley on the Hertfordshire and Cambridge border. It accords with the general picture that the pre-Roman inhabitants, finding themselves under Roman rule so rapidly and without undue fighting, proceeded to live as before, only gradually acquiring Roman objects but only leaving their former homes in small groups. Much further field work is necessary before a good clear picture is obtained, but a concentrated effort to locate Roman sites in the river valleys with a look out for any trace of Belgic occupation will either confirm or refute this case.

Further work in excavating both hill forts and villa sites will confirm this continuity of rural life from the Iron Age to the Roman. The general principles for the conduct of agriculture had been introduced and the arrival of the Romans intensified this. With greater wealth more isolated farmhouses were built -to meet the growing demands of towns, forts and imperial taxes.